A Customized Product Recommendation with Attempting to Utilize Price Discrimination Strategy

Abstract

With the rapid development of computer technology, enterprise operation and personal life have changed profoundly. While enterprises hope to explore a better method of grasping the customer need to improve their competition, the customers expect to receive more customized services that can help them to choose the commodity easily. The emergence of the product recommendation provides a platform for resolving the two aspects of question. However, whether or not a recommendation method can bring favorable results is subject to the accuracy and customization of its process. For more exact results of recommendation, the recommendation method which this paper proposed classifies the customers into different groups before exploring the transaction records. Accurate recommendation results are expected according to the customer segmentation and the analysis on different customer groups. We also consider that the enterprises and retailers usually conduct price discrimination to meet different customers' willingness to pay for certain commodity, such as disposal, discount or bundle sales. An experiment was done based on customers' attributes and transaction records of a regional retailer in Korea, and expected results were obtained.